pz3P6gL6b/Oj8NdJKl7PTA==2025-08-03T13:00:59Zsummer 2025
I was for sure nervous coming into this course a year after hearing many bad things in Summer 2024. Overall, the revamped version of this course has led to slight improvements. I am pleased to get an A (93%) but there are still issues that bugged me and fellow peers. Please take note that the comprehensive review (pros and cons) I provide below is for a summer course, so your experience might vary in a Fall/Spring semester.
PROS:
-
The course content is spectacular. They really go into the depths of regression modeling and you learn way more than just what a typical regression equation looks like. The modules covered include simple linear regression, multiple regression, GLMs and variable selection. If you are a junior or inexperienced data professional like me, I would strongly recommend this course because it goes beyond the regression fundamentals covered in 6501 and 6203.
-
The supplemental coding files on GitHub are top notch. They give you the freedom to use either R or Python for HWs, exams and project. Regardless of your choice of language, the sample code and markdown cells provide great explanation of the deep analysis you learn to conduct in this course. Ultimate resource to use when preparing for the coding midterm. I have certainly leveled up my R game involving EDA visuals, training, testing and evaluating models thanks to this course.
-
The project is a good opportunity to implement all the technical details you learned in the class. You also get to improve your data joining/merging skills as a requirement is to find multiple datasets from various sources.
-
Exams (both theoretical and coding part) are open-book. So you can employ the CSE 6040 strategy of storing all your resources in a local folder and refer to them during the exams. No internet browsing allowed, only Stack Overflow is permitted.
CONS:
-
Be prepared for intensive self-study. I think this applies for majority of OMSA courses. The lecture videos like always are of no use. I got an A without watching a single video and instead relied on the course transcripts. So in a way, I had to learn and understand the content on my own with minimal support. You can always attend OH and/or ask questions on Piazza but I personally did not because of timing issues and cluttered Piazza threads.
-
Technical problems and errors from the teaching staff. This was a constant frustration throughout the semester. Examples of errors include posting incorrect solutions forcing students to redo peer reviews, switching to Jupyter environment for coding and providing no support for students wishing to use other IDEs, and finally messy project peer reviews that ultimately got scrapped.
-
Ambiguous wording in exams and homeworks. Another complaint raised in previous semesters that has not been addressed. I once read somewhere that since the exams are switched to open book, this is a tradeoff we simply have to accept.
GENERAL TIPS TO SUCCEED:
-
Don't waste time on watching lecture videos. Download/print the lecture transcripts. Read them, highlight them, understand them.
-
Reviewing stats before taking the course will make the understanding of course content smoother as there are many references to estimation. confidence intervals, distributions, degrees of freedom, etc. Calculus and linear algebra not needed.
-
Ahead of the exams, make organized folders where you can store your files. For the MCQ portion, transcripts, personal notes, printed PDF of HW quizzes would be helpful. For the coding exam, the supplemental GitHub files and HW solutions would be ideal resources.
-
For the project, find group members that have experience in working with big datasets, joining data, writing professional reports and APA citations. You'll thank me later.
Rating: 3 / 5Difficulty: 3 / 5Workload: 18 hours / week